arXiv subjects

Subhajit Paul

Publications and source records attributed to Subhajit Paul.

At least 19 recordsLinked to original sources

arXiv2026

Charge Regulated conformational properties of polyelectrolyte near an oppositely charged nanoparticle

Customizing the surface characteristics and stimuli-responsive behavior of nanoparticles with polyelectrolytes ushers in a new era across many aspects of our lives, ranging from advanced diagnostics to practical applications. Here, using hybrid CR Monte Carlo/ molecular dynamics simulations, we investigate how charge regulation can play a crucial role in shaping the adsorption dynamics of polyelectrolyte (PE) on oppositely charged nanoparticle (NP). We systematically investigate the influence of salt density, and polymer chain length on the PE-NP interaction. To complement CR results, we also perform molecular simulations under constant charge conditions. At high salt concentrations, CR enhances the adsorption of PE onto the NP surface, leading to a rapid decrease in the radius of gyration of PE; conversely, CC promotes the extended conformation of PE. No clear effect of PE length is observed at either low or high salt concentrations, whereas in CC simulations, the PE relaxes faster on the NP in the case of short chains. Furthermore, By comparing these results, we demonstrate that the MSD of PE follows a more direct path during adsorption implying a ballistic motion, whereas in the CC case, it exhibits subdiffusive behavior and delayed adsorption in both low and high salt density. Our findings indicates that 'tunable CR' is a robust strategy for controlling nanoparticle stability and interaction within a complex biochemical cues.

cond-mat.soft↗

arXiv2026

Emergence of Dynamical Anisotropy induced by Demixing in a Binary System with Differential Diffusivity under an External Potential

Spontaneous demixing in active matter is a ubiquitous phenomenon that is crucial for numerous living processes ranging from bacterial swarming to sorting of cells in dense tissues. Here, we systematically investigate the effect of spatially varying potential acting along one direction and packing fraction on the binary mixture of particles with different diffusivities. Our results indicate that the presence of an external potential promotes demixing over a larger range of packing fractions, while also fostering a more pronounced 'hexatic order' within the bands of less diffusive "cold") particles formed near the minima of the potential. The mean-squared displacements (MSD) of "cold" and "hot" particles in different directions exhibit a distinct behavior. In contrast to the long-time sub-diffusive behavior of the "cold" particles, the "hot" ones display diffusive nature following an intermediate plateau. However, in the direction transverse to the applied potential, both types of particles undergo normal diffusion. Furthermore, interesting non-Gaussian characteristics are observed, corresponding to the spatial distribution of the displacement of "hot" and "cold" particles. Interestingly, our results reveal the formation of a 'percolating band', and the emergence of such dynamic anisotropy is not observed in the absence of an external potential. These aspects are highly relevant to the dynamics of various systems-including densely packed tissues, bacterial motility in confined spaces, and granular segregation in the pharmaceutical industry.

cond-mat.soft↗

arXiv2026

From Global Flocking to Local Clustering: Interplay between Velocity Alignment and Visual Perception of Active Particles

Collective behavior in biological systems was first captured by the Vicsek model, in which particles align their velocities in the average direction of neighbors, leading to coherent motion and showing an order-disorder transition. However, in many complex environments, the interactions are non-reciprocal, lacking an action-reaction symmetry. Using framework of the Vicsek model, we implement non-reciprocity by restricting interactions to neighbors located inside a finite vision cone, for a particle by limiting its set of interacting neighbors which fall within a vision-cone, providing a minimal description for cognitive perception. Using detailed numerical simulations, we explore the clustering and flocking behavior due to competition between noise and limited visual perception in the presence of alignment interaction. For low noise, with reduction in the vision angle the system shows transition from a global coherent motion to locally ordered small-sized clusters. This behavior is confirmed through the steady-state distributions of velocity components and their fluctuation relative to the global mean. This is also characterized using a polar order-parameter and a two-point velocity correlation function. Interestingly, at small vision angles, particles exhibit strong short-range correlations within clusters even in the absence of any global coherence. Time-evolution of the related correlation functions illustrate the pathways towards the emergence of such structures. The time dependence of the average cluster size and the length-scale calculated from the two-point velocity correlation show scaling behavior and indicate that the emergence of density field clustering is a consequence of the velocity-field coherence. Any kind of ordering and clustering disappear in the limit of high noise and low vision-angle regime.

cond-mat.soft↗

arXiv2026

Interplay of activity and non-reciprocity in tracer dynamics: From non-equilibrium fluctuation-dissipation to giant diffusion

Non-reciprocal interactions play a key role in shaping transport in active and passive systems, giving rise to striking nonequilibrium behavior. Here, we study the dynamics of a tracer -- active or passive -- embedded in a bath of active or passive particles, coupled through non-reciprocal interactions. Starting from the microscopic stochastic dynamics of the full system, we derive an overdamped generalized Langevin equation for the tracer, incorporating a non-Markovian memory kernel that captures bath-mediated correlations. This framework allows us to compute the tracer's velocity and displacement response, formulate a generalized nonequilibrium fluctuation-dissipation relation, and determine the mean-squared displacement (MSD). We find that while the MSD becomes asymptotically diffusive, the effective diffusivity depends non-monotonically on the degree of non-reciprocity and exhibits a pronounced enhancement near a resonance set by the number of bath particles. We refer to this regime as giant diffusivity. We further show that this enhanced transport is accompanied by a strong increase in heat dissipation, revealing a direct thermodynamic cost of non-reciprocal transport. Our analytical predictions are supported by numerical simulations, including systems with short-range interactions, demonstrating the robustness of giant diffusivity and highlighting experimentally accessible signatures of non-reciprocal interactions in soft materials.

cond-mat.stat-mech↗

arXiv2025

Crossover dynamics and non-Gaussian fluctuations in inertial active chains

We study the dynamics of inertial active particles in a one-dimensional chain with harmonic nearest-neighbor interactions, highlighting the interplay of persistence, interaction, and inertial timescales. Using a Green's function approach, we derive the mean-squared displacement (MSD) and mean-squared change in velocity (MSCV), revealing multiple crossovers between ballistic, diffusive, and subdiffusive regimes and providing analytic expressions for scaling coefficients and crossover times. Non-Gaussian deviations in active Brownian particles are captured through excess kurtosis, reflecting heavy-tailed, finite-support, or bimodal distributions that evolve systematically over time. Time-dependent probability distributions exhibit distinct data collapses within different temporal regimes, confirming the robustness of the scaling behavior. Overall, this framework connects multiparticle interactions to microscopic dynamics, revealing experimentally accessible signatures of inertia in active matter.

cond-mat.stat-mech↗

arXiv2025

Density-field structures in a few systems undergoing velocity ordering

We consider two (off-lattice) varieties of out-of-equilibrium systems, viz., granular and active matter systems, that, in addition to displaying velocity ordering, exhibit fascinating pattern formation in the density field, similar to those during vapor-liquid phase transitions. In the granular system, velocity ordering occurs due to reduction in the normal components of velocities, arising from inelastic collisions. In the active matter case, on the other hand, velocity alignment occurs because of the inherent tendency of the active particles to follow each other. Inspite of this difference, the patterns, even during density-field evolutions, in these systems can be remarkably similar. This we have quantified via the calculations of the two-point equal time correlation functions and the structure factors. These results have been compared with the well studied case of kinetics of phase separation within the framework of the Ising model. Despite the order-parameter conservation constraint in all the cases, in the density field, the quantitative structural features in the Ising case is quite different from those for the granular and active matters. Interestingly, the correlation function for the latter varieties, particularly for an active matter model, quite accurately describes the structure in a real assembly of biologically active particles.

cond-mat.soft↗

arXiv2024

Anomalous Conformations and Dynamics of Active Block Copolymers

Heterogeneous distribution of passive and active domains in the chromosome plays a crucial role for its dynamic organization within the cell nucleus. Motivated by that here we investigate the steady-state conformation and dynamics of a model active-block copolymer using numerical simulations. Our results show that depending on the relative arrangements of the active and passive blocks, the polymer shows an unusual swelling, even larger than the corresponding fully active polymer. On the one hand, the dynamics of the full polymer show usual enhanced diffusion and Rouse-like scaling behavior. On the other hand, individual passive and active blocks show anomalous transient super- and sub-diffusive dynamics. We characterize this anomalous dynamics in terms of the dependence of a generalized diffusion constant with the polymer length and activity strength.

cond-mat.soft↗

arXiv2024

A Sinkhorn Regularized Adversarial Network for Image Guided DEM Super-resolution using Frequency Selective Hybrid Graph Transformer

Digital Elevation Model (DEM) is an essential aspect in the remote sensing (RS) domain to analyze various applications related to surface elevations. Here, we address the generation of high-resolution (HR) DEMs using HR multi-spectral (MX) satellite imagery as a guide by introducing a novel hybrid transformer model consisting of Densely connected Multi-Residual Block (DMRB) and multi-headed Frequency Selective Graph Attention (M-FSGA). To promptly regulate this process, we utilize the notion of discriminator spatial maps as the conditional attention to the MX guide. Further, we present a novel adversarial objective related to optimizing Sinkhorn distance with classical GAN. In this regard, we provide both theoretical and empirical substantiation of better performance in terms of vanishing gradient issues and numerical convergence. Based on our experiments on 4 different DEM datasets, we demonstrate both qualitative and quantitative comparisons with available baseline methods and show that the performance of our proposed model is superior to others with sharper details and minimal errors.

eess.IV↗

arXiv2024

F2former: When Fractional Fourier Meets Deep Wiener Deconvolution and Selective Frequency Transformer for Image Deblurring

Recent progress in image deblurring techniques focuses mainly on operating in both frequency and spatial domains using the Fourier transform (FT) properties. However, their performance is limited due to the dependency of FT on stationary signals and its lack of capability to extract spatial-frequency properties. In this paper, we propose a novel approach based on the Fractional Fourier Transform (FRFT), a unified spatial-frequency representation leveraging both spatial and frequency components simultaneously, making it ideal for processing non-stationary signals like images. Specifically, we introduce a Fractional Fourier Transformer (F2former), where we combine the classical fractional Fourier based Wiener deconvolution (F2WD) as well as a multi-branch encoder-decoder transformer based on a new fractional frequency aware transformer block (F2TB). We design F2TB consisting of a fractional frequency aware self-attention (F2SA) to estimate element-wise product attention based on important frequency components and a novel feed-forward network based on frequency division multiplexing (FM-FFN) to refine high and low frequency features separately for efficient latent clear image restoration. Experimental results for the cases of both motion deblurring as well as defocus deblurring show that the performance of our proposed method is superior to other state-of-the-art (SOTA) approaches.

cs.CV↗

arXiv2024

Remembering Everything Makes You Vulnerable: A Limelight on Machine Unlearning for Personalized Healthcare Sector

As the prevalence of data-driven technologies in healthcare continues to rise, concerns regarding data privacy and security become increasingly paramount. This thesis aims to address the vulnerability of personalized healthcare models, particularly in the context of ECG monitoring, to adversarial attacks that compromise patient privacy. We propose an approach termed "Machine Unlearning" to mitigate the impact of exposed data points on machine learning models, thereby enhancing model robustness against adversarial attacks while preserving individual privacy. Specifically, we investigate the efficacy of Machine Unlearning in the context of personalized ECG monitoring, utilizing a dataset of clinical ECG recordings. Our methodology involves training a deep neural classifier on ECG data and fine-tuning the model for individual patients. We demonstrate the susceptibility of fine-tuned models to adversarial attacks, such as the Fast Gradient Sign Method (FGSM), which can exploit additional data points in personalized models. To address this vulnerability, we propose a Machine Unlearning algorithm that selectively removes sensitive data points from fine-tuned models, effectively enhancing model resilience against adversarial manipulation. Experimental results demonstrate the effectiveness of our approach in mitigating the impact of adversarial attacks while maintaining the pre-trained model accuracy.

cs.LG↗

arXiv2024

Dynamical crossovers and correlations in a harmonic chain of active particles

We explore the dynamics of a tracer in an active particle harmonic chain, investigating the influence of interactions. Our analysis involves calculating mean-squared displacements (MSD) and space-time correlations through Green's function techniques and numerical simulations. Depending on chain characteristics, i.e., different time scales determined by interaction stiffness and persistence of activity, tagged-particle MSD exhibit ballistic, diffusive, and single-file diffusion (SFD) scaling over time, with crossovers explained by our analytic expressions. Our results reveal transitions in bulk particle displacement distributions from an early-time bimodal to late-time Gaussian, passing through regimes of unimodal distributions with finite support and negative excess kurtosis and longer-tailed distributions with positive excess kurtosis. The distributions exhibit data collapse, aligning with ballistic, diffusive, and SFD scaling in the appropriate time regimes. However, at much longer times, the distributions become Gaussian. Finally, we derive expressions for steady-state static and dynamic two-point displacement correlations, consistent with simulations and converging to equilibrium results for small persistence. Additionally, the two-time stretch correlation extends to longer separation at later times, while the autocorrelation for the bulk particle shows diffusive scaling beyond the persistence time.

cond-mat.stat-mech↗

arXiv2024

Thermodynamically Stable Knots in Semiflexible Polymers

Semiflexible polymers are widely used as a paradigm for understanding structural phases in biomolecules including folding of proteins. Here, we compare bead-spring and bead-stick variants of coarse-grained semiflexible polymer models that cover the whole range from flexible to stiff by conducting extensive replica-exchange Monte Carlo computer simulations. In the data analysis we focus on knotted conformations whose stability is shown to depend on the ratio $r_b/r_{\rm min}$ with $r_b$ denoting the equilibrium bond length and $r_{\rm min}$ the distance of the strongest nonbonded interactions. For both models, our results provide evidence that at low temperatures for $r_b/r_{\rm min}$ outside a small range around unity one always encounters knots as generic stable phases along with the usual frozen and bent-like structures. By varying the bending stiffness, we observe rather strong first-order-like structural transitions between the coexisting phases characterized by these geometrically different motifs. Through analyses of the energy distributions close to the transition point, we present exploratory estimates of the free-energy barriers between the coexisting phases.

cond-mat.soft↗

arXiv2024

Spontaneous Micro Flocking of Active Inertial Particles without Alignment Interaction

Observing spontaneous velocity ordering or flocking during motility induced phase separation (MIPS) in a system of spherical active Brownian particles without alignment interaction is challenging. We take up this problem by performing simulations of spherical active inertial particles with purely repulsive potential in presence of thermal noise and absence of any explicit alignment interaction. Our results not only show the presence of MIPS, but also reveal a micro-flocking transition. We characterize this transition in terms of a velocity order parameter as well as a characteristic length scale derived from the spatial correlation of the velocities.

cond-mat.soft↗

arXiv2023

CLiSA: A Hierarchical Hybrid Transformer Model using Orthogonal Cross Attention for Satellite Image Cloud Segmentation

Clouds in optical satellite images are a major concern since their presence hinders the ability to carry accurate analysis as well as processing. Presence of clouds also affects the image tasking schedule and results in wastage of valuable storage space on ground as well as space-based systems. Due to these reasons, deriving accurate cloud masks from optical remote-sensing images is an important task. Traditional methods such as threshold-based, spatial filtering for cloud detection in satellite images suffer from lack of accuracy. In recent years, deep learning algorithms have emerged as a promising approach to solve image segmentation problems as it allows pixel-level classification and semantic-level segmentation. In this paper, we introduce a deep-learning model based on hybrid transformer architecture for effective cloud mask generation named CLiSA - Cloud segmentation via Lipschitz Stable Attention network. In this context, we propose an concept of orthogonal self-attention combined with hierarchical cross attention model, and we validate its Lipschitz stability theoretically and empirically. We design the whole setup under adversarial setting in presence of Lovász-Softmax loss. We demonstrate both qualitative and quantitative outcomes for multiple satellite image datasets including Landsat-8, Sentinel-2, and Cartosat-2s. Performing comparative study we show that our model performs preferably against other state-of-the-art methods and also provides better generalization in precise cloud extraction from satellite multi-spectral (MX) images. We also showcase different ablation studies to endorse our choices corresponding to different architectural elements and objective functions.

cs.CV↗

arXiv2023

High-resolution Multi-spectral Image Guided DEM Super-resolution using Sinkhorn Regularized Adversarial Network

Digital Elevation Model (DEM) is an essential aspect in the remote sensing domain to analyze and explore different applications related to surface elevation information. In this study, we intend to address the generation of high-resolution DEMs using high-resolution multi-spectral (MX) satellite imagery by incorporating adversarial learning. To promptly regulate this process, we utilize the notion of polarized self-attention of discriminator spatial maps as well as introduce a Densely connected Multi-Residual Block (DMRB) module to assist in efficient gradient flow. Further, we present an objective function related to optimizing Sinkhorn distance with traditional GAN to improve the stability of adversarial learning. In this regard, we provide both theoretical and empirical substantiation of better performance in terms of vanishing gradient issues and numerical convergence. We demonstrate both qualitative and quantitative outcomes with available state-of-the-art methods. Based on our experiments on DEM datasets of Shuttle Radar Topographic Mission (SRTM) and Cartosat-1, we show that the proposed model performs preferably against other learning-based state-of-the-art methods. We also generate and visualize several high-resolution DEMs covering terrains with diverse signatures to show the performance of our model.

eess.IV↗

arXiv2023

Activity Induced Enhanced Diffusion of a Polymer in Poor Solvent

By means of Brownian dynamics simulations we study the steady-state dynamic properties of a flexible active polymer in a poor solvent condition. Our results show that the effective diffusion constant of the polymer $D_{\rm eff}$ gets significantly enhanced as activity increases, much like in active particles. The simulation data are in agreement with a theoretically constructed Rouse model of active polymer, demonstrating that irrespective of the strength of activity, the long-time dynamics of the polymer chain is characterized by a universal Rouse-like scaling $D_{\rm eff} \sim N^{-1}$, where $N$ is the chain length.

cond-mat.soft↗

arXiv2022

Activity Mediated Globule to Coil Transition of a Flexible Polymer in Poor Solvent

Understanding the role of self-propulsion on the conformational properties of active filamentous objects has relevance in biology. In this context, we consider a flexible bead-spring model polymer for which along with both attractive and repulsive interactions among the non-bonded monomers, activity for each bead works along its intrinsic direction of self-propulsion. We study its kinetics in the overdamped limit, following a quench from good to poor solvent condition. We observe that with low activities, though the kinetic pathways remain similar, the scaling exponent for the relaxation time of globule formation becomes smaller than that for the passive case. Interestingly, for higher activities when self-propulsion dominates over interaction energy, the polymer becomes more extended. In its steady state, the variation of the spatial extension of the polymer, measured via its gyration radius, shows two completely different scaling regimes: The corresponding Flory exponent changes from $1/3$ to $3/5$ similar to a transition of the polymer from a globular state to a self-avoiding walk. This can be explained by an interplay among three energy scales present in the system, viz., the "ballistic", thermal, and interaction energy.

cond-mat.soft↗

arXiv2022

Role of temperature and alignment activity on kinetics of coil-globule transition of a flexible polymer

We study the nonequilibrium kinetics during the coil-globule transition of a flexible polymer chain with active beads after a quench from good to poor solvent condition using molecular dynamics simulation. Activity for each bead is introduced via the well-known Vicsek-like alignment rule due to which the velocity of a bead tries to align towards the average direction of its neighbors. We investigate the role of quenching temperature with varying activity during collapse of this polymer. We find that although for lower activities the kinetics remains qualitatively similar for different temperatures, for higher activity noticeable differences can be identified.

cond-mat.soft↗